THE ROLE OF MYCORRHIZAE IN MEDITERRANEAN ECOSYSTEM REVEGETATION. Patrícia Maria Ferreira Correia

UNIVERSIDADE DE LISBOA FACULDADE DE CIÊNCIAS DEPARTAMENTO DE BIOLOGIA VEGETAL THE ROLE OF MYCORRHIZAE IN MEDITERRANEAN ECOSYSTEM REVEGETATION Patrícia Maria Ferreira Correia DOUTORAMENTO EM BIOLOGIA (ECOLOGIA) 2006

UNIVERSIDADE DE LISBOA FACULDADE DE CIÊNCIAS DEPARTAMENTO DE BIOLOGIA VEGETAL THE ROLE OF MYCORRHIZAE IN MEDITERRANEAN ECOSYSTEM REVEGETATION Patrícia Maria Ferreira Correia Tese orientada por: Professora Maria Amélia Botelho Paulo Martins Campos Loução Professor John Klironomos DOUTORAMENTO EM BIOLOGIA (ECOLOGIA) 2006

DECLARAÇÃO O trabalho apresentado nesta dissertação foi desenvolvido no Centro de Ecologia e Biologia Vegetal (Faculdade de Ciências da Universidade de Lisboa, Portugal) e no Department of Integrative Biology (University of Guelph, ON, Canadá). O presente trabalho foi financiado pela Fundação para a Ciência e Tecnologia, através da bolsa PRAXIS XXI/BD/15918/98. Os resultados apresentados nesta dissertação foram incluídos nos seguintes manuscritos: Patrícia Correia, Luís Carvalho, Alice Tavares, M. Amélia Martins-Loução, John Klironomos (2004). Using Native Plants to Assess Arbuscular Mycorrhizal Fungi When Restoring Quarries to Maquis Ecosystems (Portugal). Ecological Restoration 22 (3): 233-234. Patrícia Correia, Luís Carvalho, Alice Tavares, M. Amélia Martins-Loução, John Klironomos. Using indigenous and commercial arbuscular mycorrhizal fungi to grow native plants for Mediterranean ecosystem restoration (em preparação) De acordo com o disposto no n o 2, do artigo 8, do Decreto-Lei 388/70, a autora desta dissertação declara que interveio no planeamento e execução do trabalho experimental, na interpretação dos resultados obtidos e na elaboração dos manuscritos. 26 de Abril de 2006 Patrícia Maria Ferreira Correia

AGRADECIMENTOS & ACKNOWLEDGMENTS A dear friend of mine mentioned that probably the acknowledgments of my thesis would be the largest chapter, due to my continual request for help in the field and in the laboratory. It was not (!), but I feel blessed and fortunate for having so many wonderful persons sharing this endeavour with me. À Professora Doutora Maria Amélia Martins-Loução por me ter entusiasmado e possibilitado financeiramente o desenvolvimento do meu interesse pelas associações micorrízicas, bem como os primeiros ensinamentos nesta área de estudo, muito antes dos trabalhos de doutoramento. Agradeçolhe ainda por me envolver em projectos internacionais permitindo contactar outros investigadores desta área e ultrapassando o meu isolamento inicial, por sempre me facilitar viagens a laboratórios estrangeiros para aprendizagem de técnicas novas, ou participação em reuniões científicas. Agradeço-lhe ainda a disponibilidade para resolver as inúmeras dificuldades administrativas que surgiram ao longo deste período de trabalho. I am very grateful to Doctor John Klironomos for believing in my enthusiastic first email, giving me the opportunity to work in his laboratory for several periods along four years. By introducing me to new scientific approaches and mycorrhizal ecological concepts, he had a major influence in the delineation of this thesis. His foresight and constructive attitude as a scientist were a major motivation for my research. My family and I are extremely grateful for his unconditional financial support by shipping all my personal belongings from Canada to Portugal, and for his donation to cancer research on my father s name. I am deeply indebted to Professor Richard Reader (head of Botanical Department of the University of Guelph), for being co-advisor of my thesis for 6 months, sparing his time, effort and patience in teaching me fundamental information on how to organize and explore my data to become scientific information. I am particularly thankful for his awareness of a personal stressed period of my life and for encouraging me to search for solutions. Ao Luís e Inês Carvalho, pela forte e sincera amizade que nos manteve sempre próximos ao longo destes quase 20 anos. Todos os agradecimentos que aqui lhes fizer serão sempre pequenos relativamente ao apoio, incentivo e carinho que incondicionalmente me disponibilizaram, bem como a generosidade com que sempre me acolheram em sua casa para longos serões micorrízicos. Agradeço em particular pela imensa ajuda prestada desde o trabalho de campo e laboratório até à revisão integral do manuscrito desta tese. Sem eles este trabalho improvavelmente chegaria a este termo. 1

À Professora Otília Correia pelos ensinamentos sobre ecologia e posteriormente sobre revegetação, e por me ter ensinado sem palavras o enorme apreço pela vegetação mediterrânica da Serra da Arrábida. Agradeço ainda o ter acreditado nas minhas ideias para conjuntamente as transformar num projecto científico na SECIL. Obrigada pelo entusiasmo e amizade, e pela constante preocupação com o desenvolvimento e término desta tese. À Professora Cristina Cruz, por me ter iniciado nos estudos de alfarrobeira e por posteriormente partilhar o meu interesse sobre associações micorrízicas, tornando-se uma força motora para a conclusão desta tese. Agradeço com amizade os bons momentos de convívio e boa disposição que passámos juntas em Portugal e em Espanha. À Professora Margarida Barata, os ensinamentos como professora de Micologia durante a licenciatura criando em mim um enorme interesse por esta área, por sempre disponibilizar o seu microscópio, bem como posteriormente a ajuda no isolamento e identificação de fungos endófitos das raízes e ainda na revisão a eles referentes no manuscrito. Agradeço profundamente a amizade que sempre me ofereceu, conjuntamente com a incondicional disponibilidade em me receber que em inúmeras situações me serviu de reconforto e alento para continuar. À Professora Teresa Gonçalves da Universidade de Coimbra, por no início dos meus estudos em micorrizas me ter ensinado com enorme simpatia e disponibilidade os seus conhecimentos nesta área e facilitando-me o acesso a literatura. O meu reconhecimento pela estadia oferecida em sua casa enquanto me ensinou várias técnicas importantes ao inicio dos estudos e por sempre ter partilhado comigo cursos intensivos em Espanha (Santiago de Compostela e Granada). À Dra. Isabel Brito, da Universidade de Évora, por no início dos meus estudos em micorrizas me ter recebido no seu laboratório e ensinado algumas técnicas na área das micorrizas, além de me facilitar o acesso a literatura. To Doctor Chris Walker my deepest gratitude for the long time he spent with my work. Without his help in identifying the AMF spores it would have been impossible to accomplish the evaluation of arbuscular mycorrhizal species diversity. À Adelaide e à Graça pela amizade que nasceu neste laboratório mas ultrapassou os limites da relação profissional, por terem acreditado nas minhas capacidades mesmo quando eu já duvidava, por conjuntamente reunirem esforços inumeráveis para concretização final do manuscrito da tese, nomeadamente à Graça nas revisões e na formatação do mesmo. É com apreço que relembro frequentemente os vários conselhos quer científicos, quer de organização pessoal e cientifica que muito me ajudaram. Um obrigada em particular à Adelaide por pacientemente me esclarecer as muitas dúvidas estatísticas, desde o desenho experimental ao tratamento de dados.

Aos meus colegas e amigos Alice Tavares, Alice Nunes, Ana Catarina, Ana Corrêa, Ana Luisa, Ana Júlia, Herculana, Luis Ferreira e Sérgio Chozas colegas e amigos que tornaram possível, fazível e aprazível este trabalho. As recordações deste período da minha vida são preenchidas com os inúmeros momentos partilhados com cada um individualmente ou em conjunto, pelos momentos de alegria no laboratório e no campo ou mesmo em jantares pirosos em casa de cada um, pelo apoio mútuo nas horas difíceis, por uma vida diariamente partilhada. Com eles enriqueci a minha vivência. Agradeço a cada um particular pelo tudo que me deram. Queria agradecer em particular: à Alice Tavares, por ter partilhado o início das grandes experiências sempre com uma determinação e energia contagiantes; à Ana Catarina, por tão frequentemente (e alegremente) me ter ajudado no laboratório e nas estufas em Setúbal; à Herculana, pelos úteis ensinamentos, bem como os inúmeros trabalhos de laboratório em que me ajudou ao longo deste anos; ao Luís e ao Sérgio pelo enorme esforço físico que fizeram nos trabalhos de recolha e transporte dos solos para o último piso da faculdade! Ao Pedro Pinho e à Patricia Silva o meu agradecimento pela indispensável e simpática ajuda (mesmo fora de horas) na análise estatística de um dos capítulos desta tese. My grateful thanks to all the members of FASEL laboratory (1999-2003) for their emotional and scientific support and for the good times we had in the laboratory that made me feel one of them! I would like to thank in particular to Vanessa, Tanya, Ben, Jen, Jeff and Theresa who always were supportive with their friendship in very difficult times of my life. Thanks for packing all my stuff and for the little presents you put in there to cheer me up! Those jazz FASEL CDs where very appreciated and often listened during the manuscript writing. Even after I left the contact remained, by precious and caring emails especially with Theresa and Vanessa and by long scientific and statistical discussion with Jeff. Aos meus amigos e colegas portugueses no Canadá, Filipe, Pedro Encarnação e Pedro Antunes, quero agradecer com especial carinho pelo seu contínuo encorajamento para terminar esta interminável tese, e particularmente ao Pedro Antunes por revisões de Inglês do manuscrito. À SECIL por ter disponibilizado meios humanos e técnicos para o fornecimento dos solos usados na parte experimental desta tese. Agradeço também ao Eng. Almeida Barbosa pela constante disponibilidade e à D. Eulália por todo o carinho com que sempre me recebeu e ensinou técnicas de viveiros. À CPPE pela cedência das instalações (estufas) e água para rega das plantas usadas neste trabalho. Agradeço em particular à Eng. Fátima Rodrigues os ensinamentos sobre germinação de sementes das plantas em estudo, bem como o fornecimento de sementes, e à D. Fernanda pelas sugestões feitas relativamente ao desenvolvimento prático dos trabalhos e a o controlo da rega. 3

À Fundação para Ciência e Tecnologia pelo financiamento da bolsa de doutoramento BD/ 15918/98 e pelo finaciamento de congressos e pagamento de deslocações e bench-fees durante estágios na universidade de Guelph (Canadá). À minha querida irmã Raquel quero agradecer o carinho com que sempre me hospedou em sua casa, por ultrapassar comigo tantos momentos difíceis, as nossas longas conversas e os seus conselhos com uma clarividência muito prática, que sempre me orientam na minha natural desorientação. Sem esta constante partilha a minha vida não fazia sentido. Aos meus queridos amigos, padrinhos e família que me apoiaram e perdoaram a minha ausência e muitas promessas não cumpridas, que aturaram as minhas crises existenciais, e ainda à Cristina e Rui por se voluntariarem em me ajudar nos trabalhos de campo ou laboratório. Àos meus pais pelo amor incondicional que sempre me deram, pelo seu apoio emocional e financeiro durante esta fase tão conturbada da minha vida. Estou eternamente reconhecida pela constante presença e ajuda, mesmo quando estive ausente do país sempre me fizeram sentir acompanhada. Agradeço-lhes por todas as ajudas que me dispensaram durante os trabalhos práticos desta tese, ao frequentemente me acompanharem às estufas da CCPE em Setúbal, pelas noitadas a preparar todo o material necessário, imenso pelo esforço físico que fizeram durante as campanhas de campo. Esta tese vos dedico por sempre terem acreditado nas minhas capacidades, mesmo quando eu duvidei. Ao meu pai o meu pedido de desculpas por não lhe ter dado esta alegria ainda em vida. I hope my father soul will rest in peace now, my promise is fulfilled!

TABLE OF CONTENTS SUMMARY...9 RESUMO...11 CHAPTER 1 - GENERAL INTRODUCTION...15 1. MYCORRHIZAE AN OVERVIEW...15 1.1. Definition...15 1.2. Arbuscular Mycorrhizal Fungi...16 1.2.1. Taxonomy...16 1.2.2. Mycorrhizal symbiosis functioning...18 1.2.3. Effect on soil structure...20 1.3. Cost-benefit of the symbiosis...21 1.3.1. Plant mycorrhizal dependence vs. plant responsiveness...21 1.3.2. Mycorrhiza association - from mutualism to parasitism...22 1.3.3. Preferential associations between host plant-amf species...23 1.4. Role of mycorrhizae in maintaining plant community diversity, linking above and below ground...24 1.4.1. Effect of AMF communities on plant communities...24 1.4.2. Effect of host plants on AMF communities...25 1.4.3. Host plant-amf interactions: feedback dynamics relation...26 1.5. Effect of disturbances on mycorrhizae, particularly in Mediterranean ecosystems...27 1.6. Mycorrhizal role in the restoration of degraded Mediterranean ecosystems...29 2. OBJECTIVES AND RATIONALE OF THE RESEARCH...33 3. EXPERIMENTAL FRAMEWORK...37 CHAPTER 2 - ASSESSING ARBUSCULAR MYCORRHIZAL FUNGI INFECTIVITY AND EFFECTIVITY PRIOR TO QUARRY RESTORATION...43 1. INTRODUCTION...43 2. MATERIALS AND METHODS...44 3. RESULTS...48 4. DISCUSSION...51 CHAPTER 3 - USING INDIGENOUS AND COMMERCIAL ARBUSCULAR MYCORRHIZAL FUNGI TO GROW NATIVE PLANTS FOR MEDITERRANEAN ECOSYSTEM RESTORATION...55 1. INTRODUCTION...55 2. MATERIALS AND METHODS...58 3. RESULTS...63 3.1. AMF diversity survey...63 3.2. AMF effectivity on plant growth...65 3.3. Root colonization...67 3.4. Plant Biomass...69 3.5. Plant effect on AMF community growth...69 4. DISCUSSION...72 CHAPTER 4 - DOES AMF SOIL COMMUNITY MEDIATE PLANT-ROOT ENDOPHYTIC FUNGI INTERACTIONS?...79 1. INTRODUCTION...79 2. MATERIALS AND METHODS...82 3. RESULTS...90 3.1. Plant Biomass...90 3.2. Root Colonization...91 3.3. Characterization of fungal communities...93 3.3.1. Mycorrhizal communities (AMF)...93 3.3.2. Non-mycorrhizal endophytic fungal community...97 3.4. Relation among fungal communities and myrtle plant biomass...103 4. DISCUSSION...105 5

CHAPTER 5 GENERAL DISCUSSION... 109 CHAPTER 6 - FINAL CONCLUSIONS... 119 CHAPTER 7 - FUTURE PERSPECTIVES... 121 CHAPTER 8 - REFERENCES... 123

LIST OF TABLES Table 1.1. Potential modes of action of AMF involved in disease biocontrol.... 20 Table 1.2. Characteristics of the plant species used in the present study... 40 Table 2.1. Soil chemical analyses..... 45 Table 3.1. Chemical characteristics of the soil mixtures before use in this assay... 60 Table 3.2. Indigenous AMF community composition from disturbed soil (early-seral AMF community) and undisturbed soil (late-seral AMF community)... 64 Table 3.3. F-values with significance levels are given for repeated measure analyses of variance (ANOVAR) of the height of three target plant species grown in soils with three different AMF communities... 65 Table 3.4. F-values with significance levels are given for three-way ANOVA of AMF soil community, plant species and harvest time effects on root colonization, plant biomass and soil AMF infectivity of three target plant species grown in soils with three different AMF communities... 67 Table 4.1. List of AMF species and their relative density (%)... 95 Table 4.2. Renkonen index values evaluating the similarity of AMF species compositions in myrtle plant rhizosphere between pairs of soil communities pre-cultured by three host plant species (myrtle, lavender and carob).... 96 Table 4.3. Renkonen index values evaluating the similarity of AMF species compositions in myrtle plant rhizosphere between pairs of soil origin (undisturbed, disturbed and disturbed soil plus commercial inoculum).... 96 Table 4.4. Ecological characteristics of the non mycorrhizal fungal genera found in second generation myrtle roots.... 99 Table 4.5. Renkonen index values evaluating the similarity of non mycorrhizal fungal endophytic communities in myrtle rhizosphere between pairs of soil communities pre-cultured by three host plant species (myrtle, lavender and carob)... 100 Table 4.6. Renkonen index values evaluating the similarity of non mycorrhizal fungal endophytic communities in myrtle rhizosphere between pairs of soil origin (undisturbed, disturbed and disturbed soil plus commercial inoculum)... 100 Table 4.7. Relative frequency (%) of the occurrence of myrtle non mycorrhizal endophytic fungi grown in the three different soil communities (undisturbed, disturbed and disturbed plus commercial inoculum).... 102 Table 4.8. Spearman rank correlations coefficients between myrtle plant biomass, root endophytic fungal colonization and fungal species diversity (mycorrhizal and non mycorrhizal) in nine different soil communities... 103 Table 4.9. Summary of the Stepwise regression analysis concerning the variation of myrtle plant biomass... 105 Table 5.1. Summary of the hypotheses proposed in Chapter 1 for the effect of disturbance, commercial inoculum addition and native plants on AMF communities... 117 7

IST OF FIGURES Figure 1.1. The diagnostic structural features of the six recognised types of mycorrhizae. 16 Figure 1.2. Mycorrhizal phenotypes are manifestations of the interaction between plant and fungal genotypes and environmental conditions...22 Figure 1.3. Feedback due to changes in composition of the community of soil mutualists.26 Figure 1.4. A schematic representation of a protocol for managing mycorrhizal fungi in soils to be revegetated....23 Figure 1.5. Schematic representation of hypothetical disturbance effect for the dependent variables...27 Figure 1.6. Schematic representation of hypothetical plant growth responsiveness to three different AMF communities...28 Figure 1.7. Aspect of the reference undisturbed site, in Serra da Arrábida (A), of the greenhouse facilities where the experiments were carried out, in Praias do Sado (B), and plant species used carob (C), myrtle (D) and lavender (E)...37 Figure 2.1. Direct count of AMF propagules (black bars) and their viability (white bars) in undisturbed and disturbed soil...47 Figure 2.2. Biomass expressed as fresh weigh (g), and root/shoot ratio for the different host plants and their respective root colonization percentage under disturbed (white bars) and undisturbed treatments (black bars)....50 Figure 3.1. Schematic representation of the experimental working hypotheses....50 Figure 3.2. Schematic representation of experimental AMF treatments set-up....53 Figure 3.3. Time-course effect of the AMF treatments on the heights of three host-plant species...66 Figure 3.4. Root length colonization at plant harvest after 6 and 12 months of three target plant species grown in soils with three different AMF communities...60 Figure 3.5. Plant biomass at plant harvest after 6 and 12 months of three target plant species grown in soils with three different AMF communities...62 Figure 3.6. Variation of AMF soil infectivity promoted by three different host-plant species, measured by Sorghum bioassay....63 Figure 4.1. Schematic diagram of hypothetical plant-root endophytic fungal feedback mediated by the soil mycorrhizal community...73 Figure 4.2. Schematic representation of the experimental set-up...77 Figure 4.3. Total plant biomass of one-year-old myrtle plants grown in each of the nine soil community treatments...91 Figure 4.4. Root length colonization by mycorrhizal (AM) and non mycorrhizal endophytic fungi of myrtle plants grown in nine different soil communities...92 Figure 4.5. AMF species diversity at harvest of the second generation of myrtle plants in response to different soil communities....93 Figure 4.6. Shanon index diversity of non mycorrhizal root endophytic fungal species in response to different soil communities....98 Figure 4.7. Representation of multiple linear regression of myrtle plant biomass with AMF predictor species abundance....96

SUMMARY The reestablishment of a functional soil microbial community, in particular arbuscular mycorrhizal fungi (AMF), is crucial for successful plant establishment in ecosystem restoration trials. AMF soil inoculation is suggested for these extreme situations. However, little is known about its beneficial effects on woody Mediterranean plants. The overall aim of this research was to provide more information concerning the addition of commercial AMF inoculum in nursery-grown plants, as an approach to overcome constraints to Mediterranean quarry restoration, and to evaluate its effects on the composition of the indigenous AMF community remaining in the soil after disturbance. Effectiveness of indigenous AMF communities from disturbed soil prior to inoculation, and from undisturbed nearby maquis soil were also assessed. Three specific questions were addressed: 1. Does disturbance act as a selective force, shifting AMF community composition by selecting less infective and less effective AMF species? 2. Do changes in species diversity and abundance of AMF communities, driven by disturbance and commercial inoculum addition, induce changes on native plant growth benefits? Are those changes plant species dependent? 3. Do changes in AMF communities, driven by disturbance and commercial inoculum addition, induce changes on plant-root endophyte interactions? Overall, the results led to the following conclusions: 1. Disturbance had a negative effect on AMF propagule density and infectivity as measured through sorghum root colonization, although this effect was not related to AMF effectivity in woody plants. 2. Adding commercial AMF inoculum or using disturbed soil containing only indigenous AMF species, were both efficient in promoting growth for all studied woody plant species. Plants were able to increase diversity and infectivity of AMF communities, although plant species-specific variations were observed. 3. Native plant-root endophytical fungal (mycorrhizal and non mycorrhizal) feedback did not affect plant growth. Plant biomass variation was associated with changes in AMF species abundance. KEYWORDS: arbuscular mycorrhizae, commercial mycorrhizal fungi inoculum, Mediterranean woody plants, restoration, soil disturbance 9

RESUMO Actualmente, a recuperação de ecossistemas degradados directa ou indirectamente pela acção humana, como por exemplo de pedreiras, é cada vez mais importante. Durante a exploração das pedreiras o solo é perturbado, sendo retirado e armazenado até ser reutilizado nas áreas a revegetar. Esta perturbação reduz a viabilidade dos microganismos do solo, e a erosão dos solos durante o armazenamento leva à perda de nutrientes. Assim, a qualidade microbiológica do solo e as espécies de plantas selecionadas para revegetar serão factores condicionantes do sucesso da revegetação. É actualmente aceite que as micorrizas arbusculares (AM) são microrganismos fundamentais na recuperação de ecossistemas perturbados, dado que estas simbioses mutualistas podem aumentar a capacidade das plantas para tolerar e sobreviver a situações adversas. Estas associações possibilitam à planta um aumento de nutrientes enquanto o fungo beneficia dos hidratos de carbono provenientes da actividade fotossintética da planta. Vários estudos sugerem a manipulação dos fungos micorrízicos em planos de revegetação. Para tal são recomendados dois tipos de avaliação sequencial: 1) estudo de infectividade dos fungos micorrízicos que possam ainda existir no solo perturbado, isto é, capacidade dos fungos para colonizarem as raízes das plantas; 2) estudo da efectividade destes fungos, isto é, da promoção do crescimento das plantas. Se ambas as avaliações forem negativas, deverá efectuar-se a adição de inóculo micorrízico externo, como estratégia para ultrapassar as limitações que a redução de fungos micorrízicos possa trazer ao desenvolvimento das plantas durante a revegetação. Esta aplicação de inóculo micorrízico poderá ser feita directamente no campo ou durante o processo de produção em viveiro das plantas a usar na revegetação. Se, pelo contrário, a comunidade de fungos micorrízicos for ainda infectiva, é então sugerida a sua manipulação indirecta através do uso de herbáceas ou de plantas de crescimento rápido, para aumentar o número de propágulos micorrízicos no solo. Ambas as estratégias se baseiam em estudos que demonstram a existência de uma relação linear entre a densidade de propágulos no solo, a colonização das raízes e o aumento do benefício da associação micorrízica para as plantas. No entanto, a investigação actual sobre as associações micorrízicas, tem demonstrado que tal linearidade não é universal. A associação micorrízica nem sempre é verdadeiramente mutualista, podendo inclusivé ser prejudicial ao crescimento das plantas, dependendo das espécies de fungos envolvidas. Não é possível, portanto, generalizar a função benéfica destes fungos para todas as espécies de plantas. O resultado da simbiose pode ainda variar com as espécies das plantas envolvidas, e com as condições abióticas e bióticas do sistema em estudo. Dada a complexidade de factores que fazem variar esta relação, torna-se difícil 11

prever o sucesso da manipulação das fungos micorrízicos para favorecem o crescimento das plantas numa revegetação. A investigação desta tese procurou avaliar a funcionalidade das associações micorrízicas estabelecidas entre diferentes comunidades fúngicas e plantas lenhosas mediterrânicas. De acordo com conhecimentos científicos recentes, a avaliação do sucesso das associações micorrízicas é conseguida através do efeito que os fungos exercem no crescimento das plantas e, reciprocamente, do efeito que as plantas exercem na comunidade nativa de fungos micorrízicos. Neste trabalho seleccionaram-se três comunidades de fungos micorrízicos: (i) comunidade existente num solo perturbado pelas actividades extractivas e de armazenamento numa pedreira; (ii) comunidade resultante da adição de inóculo micorrízico comercial à comunidade existente no solo perturbado; (iii) comunidade encontrada no solo de uma área de maquis não perturbada, próxima da pedreira, e onde naturalmente crescem as espécies de plantas estudadas. As comunidades fúngicas seleccionadas para este estudo têm, assim, relevância para possíveis manipulações de inóculo micorrízico em planos de revegetação. As plantas seleccionadas foram de espécies lenhosas mediterrânicas, pertencentes a estádios serais da sucessão diferentes e consequentemente com diferentes estratégias adaptativas: Ceratonia siliqua (alfarrobeira), Myrtus communis (murta) e Lavandula stoechas (alfazema). Pretendeu-se especificamente responder às seguintes questões: 1. Será o efeito da perturbação no solo uma força selectiva, capaz de induzir diferenças na composição da comunidade fúngica, alterando assim a infectividade e efectividade dos fungos micorrízicos? (Capítulos 2 e 3) 2. Será que diferenças nas comunidades de fungos micorrízicos, devidas à perturbação do solo e à adição de inóculo comercial, induzem diferentes benefícios no crescimento de plantas mediterrânicas? Dependerá esse efeito da espécie da planta? (Capítulo 3) 3. As diferentes comunidades de fungos micorrízicos irão alterar as interacções que existem entre as plantas e os seus fungos endófitos das raízes? (Capítulo 4) Os resultados mais relevantes obtidos em resposta a estas questões resumem-se em seguida. No Capítulo 2, o solo perturbado foi comparado com o solo não perturbado, para responder à pergunta 1. Foram utilizadas as três espécies de plantas hospedeiras acima mencionadas e, como planta-referência de crescimento rápido, foi usado o sorgo, muito comum em estudos de avaliação de infectividade de fungos micorrízicos. Os resultados mostraram que a perturbação teve um efeito negativo na densidade de propágulos e,

simultaneamente, na infectividade determinada com o sorgo, mas não com as plantas mediterrânicas. Pode-se então concluir que a perturbação altera a densidade de propágulos micorrízicos mas não a sua capacidade de colonizar plantas mediterrânicas. No Capítulo 3, e ainda para responder à pergunta 1, confirmaram-se as diferenças entre a diversidade das comunidades de fungos micorrízicos seleccionadas. De acordo com o inicialmente previsto, a comunidade de fungos micorrízicos resultante da perturbação do solo apresentou a menor densidade e diversidade de esporos, enquanto a comunidade do solo não perturbado apresentou maior diversidade. A adição de inóculo comercial induziu alterações na comunidade ao nível da densidade de esporos. Para responder à pergunta 2, o desenho experimental deste capitulo foi delineado para comparar as três comunidades fúngicas e a sua efectividade nas plantas mediterrânicas. Os resultados deste estudo não mostraram aumento do crescimento das plantas mediterrânicas devido à adição de inóculo micorrízico comercial ao solo. Pelo contrário, a comunidade de fungos do solo perturbado promoveu mais o crescimento das plantas esclerófilas (alfarrobeira e murta), comparativamente à espécie semi-decídua (alfazema), do que a comunidade de fungos do solo não perturbado. Neste estudo, as alterações da densidade de propágulos micorrízicos no solo não induziram alterações da efectividade desses fungos na promoção do crescimento das plantas. Por outro lado, as plantas mediterrânicas foram capazes de manter a infectividade do inóculo comercial ao longo de um ano, tendo a murta e a alfazema aumentado a capacidade infectiva da comunidade de fungos micorrízicos do solo não perturbado. Foi evidenciado que o benefício das diferentes plantas mediterrânicas na communidade fúngica dependeu da espécie de planta usada. No Capítulo 4 procurou-se responder à terceira pergunta. Tem sido sugerido por vários autores que o inóculo micorrízico adicionado ao solo tem a capacidade de actuar como agente bioprotector. Assim, seria esperado que o inóculo comercial adicionado ao solo mediasse as interacções entre as plantas e os fungos endófitos da raiz (micorrízicos e não micorrízicos). No entanto, a adição de Glomus intraradices não alterou a diversidade e/ou abundância dos fungos micorrízicos da comunidade do solo perturbado, nem dos fungos não micorrízicos endófitos da raiz. Pode então ser sugerido, pelo menos neste estudo, que G. intraradices não actuou directamente como agente bioprotector. Observou-se, pelo contrário, um efeito negativo da comunidade conspecífica de microrganismos do solo no crescimento de murta. Aparentemente, esta redução do crescimento das plantas de murta está inversamente associada à variação de abundância de espécies nativas de fungos micorrízicos (G. geosporum e G. constrictum) e não à presença de fungos patogénicos nas raízes. 13

Como conclusão geral pode afirmar-se que a variabilidade observada no crescimento das plantas em função das diferentes comunidades de fungos micorrízicos, e no recíproco efeito das diferentes espécies de plantas nas comunidades fúngicas, se relaciona mais com as espécies de plantas e de fungos envolvidas do que com o efeito da perturbação ou da adição de inóculo comercial micorrízico. PALAVRAS-CHAVE: inóculo de fungos micorrízicos; micorrizas arbusculares; perturbação do solo; plantas mediterrânicas; revegetação.

Chapter 1 GENERAL INTRODUCTION Chapter 1 - GENERAL INTRODUCTION 1. MYCORRHIZAE AN OVERVIEW 1.1. DEFINITION It was A.B. Frank in 1885 who used the term mycorrhiza for the first time. Since then, research progressed towards the description and characterization of the morphology and function of these organisms (see Koide and Mosse, 2004). Gerdemann (1970) defined mycorrhiza as a mutualistic symbiotic association between soil-borne fungi and plants. It is generally accepted that these symbiotic organisms have co-evolved since the first terrestrial plants appeared (Simon et al., 1993; Taylor et al., 1995; Gonçalves and Martins- Loução, 1998; Brundrett, 2002). Mycorrhizal symbioses are the most widespread among natural communities of terrestrial plants and among plant groups: in almost all liverworts and hornworts, in more than 95% of pteridophytes, in all gymnosperms and about 82% of the angiosperns (Brundrett, 2002). This association is usually considered as mutualistic (see Johnson et al., 1997), due to the highly interdependent relationship established between both partners, in which the host plant receives mineral nutrients via the fungal mycelium, and the biotrophic fungus obtains carbon compounds from the host s photosynthesis (Harley and Smith, 1983; Smith and Read, 1997). Read (2002) categorized mycorrhiza in six major types according to their morphology, structural and functional attributes related to the fungi, and plant taxa forming the symbiosis (Fig.1.1). The present study focuses on the arbuscular mycorrhiza type (AM) (re-named by Walker, 1995), the most widespread association, which will be simply referred here as mycorrhiza. In this type of mycorrhiza, the fungus develops unseptated extra- and intra-radical hyphae, intercellular arbuscules and, in some genera, vesicles. Arbuscule formation encompasses the profuse branching of hyphae after penetration of the plant cell wall. This fungal structure is closely surrounded by the plasmalemma and the large surface area thereby created between both symbionts allows the bidirectional transfer of metabolites and nutrients to and from the fungus (Gianninazzi-Pearson et al., 1996; Smith and Read, 1997). Vesicles are apical or intercalary swellings of hyphae containing lipids, which function as the energy storage organs of these fungi. The AM colonization begins with hyphae that arise from different propagules: spores, soil mycelium network and mycorrhizal root fragments (Bonfante and Peroto, 1995; Smith and Read, 1997). As 15

Chapter 1- MYCORRHIZAE AN OVERVIEW the internal colonization spreads to the youngest part of the roots, the extra-radical hyphae grow outwards into the surrounding soil. The extensive mycelium network developed will create a bridge across the nutrient depletion zone surrounding the roots and gain access to slowly mobile nutrients such as P from the bulk soil (Jakobsen, 1995). Figure 1.1. The diagnostic structural features of the six recognised types of mycorrhizae. Two basic categories are designated, one in which the root surface is sheathed in a fungal mantle (SHEATHING), and one lacking a mantle but in which hyphae proliferate internally (ENDO). The defining structures of each type are fungal pegs (MONOTROPOID), Hartig net and intracellular penetration (ARBUTOID also seen in the subtype ectendo ), Hartig net, mantle, external mycelial network (ECTO), peloton (ORCHID), hyphal complexes in hair roots (ERICOID) and arbuscles or hyphal coils (ARBUSCULAR). The most important nutrient acquisition pathways are shown. (in Martins-Loução, 2002). 1.2. ARBUSCULAR MYCORRHIZAL FUNGI 1.2.1. Taxonomy Arbuscular mycorrhizae are a monophyletic group of fungi that evolved some 400 to 460 millions of years ago, and radiated into the groups we know today as early as 350 to 400 millions of years ago (Simon et al., 1993). At the present time, and based on molecular analysis, AMF are classified as Glomeromycota, a distinct group of the Zygomycota, which were shown to be polyphyletic (Schüβler et al., 2001). Initially, the taxonomy of the group was based on the morphological features of spore walls. The choice of spores was

Chapter 1 GENERAL INTRODUCTION largely due to the obligate biotrophy of these symbionts, which cannot be cultured in the absence of their host plant (Morton and Bentivenga, 1994). Therefore, to fully evaluate AMF diversity it is necessary that the fungi achieve their reproductive phase and sporulate. This could explain the small degree of morphological diversity in comparison to the apparent high genetic diversity (Sanders, 1999). The 156 AMF species currently described are taxonomically classified into the following categories: Phylum : Glomeromycota (Schüβler et al., 2001); Order: Glomales (Morton and Benny, 1990); Suborders: Glomineae and Gigasporineae (Morton and Benny, 1990); Families: Glomaceae, Acaulosporaceae, Gigasporineae (Morton and Benny, 1990), Archaesporaceae, Paraglomaceae (Morton and Redecker, 2001) Genera: Glomus, Acaulospora, Entrophospora, Gigaspora, Scutellospora (Walker, 1987, 1992), Archaepospora and Paraglomus (Morton and Redecker, 2001) According to recent classification studies, Geosiphon sp. was added to Glomerycota. This species is non-mycorrhizal, but forms a symbiosis with cyanobacteria (Nostoc), and it is thought to be a basal lineage of the Glomales (Schüβler, 1999). There is indication that Geosiphon sp. may be less distant from characteristic Glomales than some deeply branched AMF taxa (Redecker et al., 2000), which have been placed in the new families Archaesporaceae and Paraglomaceae. Improved biochemical and molecular methods have shown that diversity may be greater in natural soils than thought until now, but the relationship between morphological and molecular diversity is still unclear (Clapp et al., 2002). The taxonomic and evolutionary relatedness between organisms is illustrated in phlylogenetic trees. The three classical families of the Glomales (Glomaceae, Acaulosporaceae and Gigasporineae) defined with morphological criteria (Morton and Benny 1990), are supported by molecular data (Simon et al., 1993). Nevertheless, there is little support for the two sub-orders Glomineae and Gigasporineae (Schüβler, 1999), and within families the relationship among species is unclear (Clapp et al., 2002). Molecular analysis (ribosomal DNA based) has proven to be advantageous comparatively to morphology-based taxonomy to consistently identify taxa present in the roots of plants (Clapp et al., 2002). In studies by Helgason et al. (1998, 2002), the dominant AMF taxa in the roots were those that were neither grown in culture nor found in the soil as spores, and 17

Chapter 1- MYCORRHIZAE AN OVERVIEW the AMF diversity was unexpectedly large. These studies suggest that many more AMF species exist than those that have already been described. Molecular studies that describe AMF communities in plant roots provide a degree of information about fungal identity, but it is unclear whether such information can be used to explain functional differences in AMF life-strategies or plant growth (van der Heijdeen, et al., 2004). This subject has been under debate. Hart and Reader (2002a,b) experimentally demonstrated that there is a link between functional traits in AMF and taxonomy at the family level. Additionally, Munkvold et al. (2004) showed that there is high functional diversity within a single AMF species with respect to P-uptake. All these drawbacks need to be clarified in order to understand the meaning of AMF diversity and functioning. Apparently, functional diversity is independent of taxonomic diversity. Spore-based characters are still the basis for Glomalean taxonomy at the species level. However, they provide little or no information on AMF life-history traits associated with abundance and architecture of fungal components, their rate of formation and longevity, and their cost to the plant host when in the symbiosis. All these features are important to the understanding at various levels, from the molecular to the ecological ones, and they are generally independent of morphological determinants (Morton and Bentivenga, 1994). On the other hand, molecular analysis has only just begun, and a streamline methodology to allowhigher throughput of samples is needed. 1.2.2. Mycorrhizal symbiosis functioning Reciprocal exchange of nutrients between symbionts is the central mechanism in mycorrhizal functioning (Jakobsen et al., 2002). Most studies of mycorrhizal functioning are made from the point of view of plant benefit from the symbiosis. AM symbiosis influences several aspects of plant physiology, such as mineral nutrition, tolerance to biotic and abiotic stresses and plant protection (Brundrett, 2002), triggering a better plant growth and development. If the plant reflects such benefits from the association, the AMF is defined as effective. The primary importance of mycorrhizae for plants results from the enlargement of the nutrient absorption surface by fungal hyphae, which enhances the roots capacity of nutrient absorption (Schüepp et al., 1987; Koide, 1991; Smith and Read, 1997). Potentially, this can maximize the acquisition of nutrients which in turn may increase primary production. Mycorrhizae can also considerably enhance nutrient mobilization,

Chapter 1 GENERAL INTRODUCTION which is particularly important for low mobile ions like phosphate and orthophosphate (Pearson and Jakobsen, 1993; Jakobsen, 1995; Smith et al., 2003). Mycorrhizae have also been shown to be important in the mobilization of: zinc, copper, iron and manganese (Ross and Harper, 1970; Gilmore, 1971; Marshner and Dell, 1994; Azaizeh et al., 1995; Liu et al., 2000); soluble inorganic nitrogen, namely nitrate and ammonium (Barea, 1991; Johansen et al., 1994; Martins-Loução et al., 2000; Yoshida and Allen, 2001, 2004; Cruz et al., 2004) and, as recently shown, organic nutrients (Koide and Kabir, 2000; Hodge et al., 2001). Other plant benefits which are not only directly associated with a better plant nutrition but also with physiological mechanisms (hormonal signaling, changes in root:shoot ratio root,) include, for example, enhanced recovery from water stress (Allen and Allen, 1986; Duan et al., 1996; Augé, 2001; Querejeta et al., 2003) and tolerance to salinity (Ruiz- Lozano and Azcón, 2000; Cantrell and Linderman, 2001; Carvalho et al., 2003) and metals (Leyval et al., 1997; Weissenhorn et al., 1994; Davies et al., 2001). The mycorrhizal symbiosis can provide protection against biotic stress, mainly to root fungal pathogens (Newsham et al., 1995; Klironomos, 2000; Borowicz, 2001). Nevertheless, some reports have shown that the symbiosis can also reduce the incidence of leaf pathogen (Lingua et al., 2002) or bacteria (Garcia-Garrido and Ocampo, 1988, 1989), and that the detrimental effects of nematodes (Van der Putten et al., 1993). The antagonistic interactions between parasitic and mycorrhizal fungi are not exerted with the same effectiveness by all AMF species. It depends on the pathogens, the plant species and the environmental conditions involved (see review by Whipps, 2004). The studies on the AMF mode of action for promoting plant protection revealed that several mechanisms can be operative for each AM fungus plant pathogen combination, and that any such interaction should be considered to be dynamic as a continuum of possible modes of action. Whipps (2004) identified several modes of action (Table 1.1). Associated microorganisms may complement the activity of AMF (Budi et al., 1999; Barea et al., 2002). The interactions of AMF with these associated soil microorganisms are not restricted to plant pathogens; the extra-radical mycelium has been shown to establish both beneficial and deleterious relations with other fungi, bacteria, nematodes and arthropods (Fitter and Garbaye, 1994; Klironomos and Ursic, 1998). Interactions between soil invertebrates and mycorrhiza seem to be an important mechanism by which these organisms structure plant communities (Klironomos and Ursic, 1998; Gange, 2001). 19

Chapter 1- MYCORRHIZAE AN OVERVIEW The AMF symbiosis changes the pattern of root exudation (Marschner et al., 1997; Pinior et al., 1999). In addition, the development of the fungal soil mycelium serves as a carbon source to soil microbial communities and introduces physical modifications into the environment surrounding the roots. These changes affect the microbial populations in the rhizosphere of the mycorrhizal plant both quantitatively and qualitatively (Barea et al., 2002; Antunes et al., 2006). Table 1.1. Potential modes of action of AMF involved in disease biocontrol (in Whipps, 2004). Reference Direct competition or inhibition Competition for photosynthate or carbon in or on root Graham 2001; Larsen and Bødker 2001; Morandi et al. 2002 Competition for exudates external to the root Schwab et al. 1984; Bansal and Mukerji 1994; Bago et al. 1996; St-Arnaud et al. 1997; Filion et al. 1999, 2003; Norman and Hooker 2000 Competition for infection sites or space on roots Liu 1995; Cordier et al. 1998; Fusconi et al. 1999; Vigo et al. 2000; Matsubara et al. 2001; Morandi et al. 2002 Quality and quantity of exudates from roots or mycorrhizal fungi inhibit Schwab et al. 1984; Bansal and Mukerji 1994; St-Arnaud et al. 1995; Filion et pathogens (potentially including low levels of antibiotics or defence compounds) al. 1999, 2003; Norman and Hooker 2000 Competitive interactions with pathogens in soil Garcia-Garrido and Ocampo 1989; St-Arnaud et al. 1994, 1995, 1997; Filion et al. 1999; Norman and Hooker 2000 Enhanced or altered plant growth, nutrition, and morphology Increased nutrient uptake (particularly phosphorus), increased trace elements, Hooker et al. 1994; Linderman 1994; Karagiannidis et al. 2002 drought tolerance, decreased toxicity to salt an heavy metals (alleviation of abiotic stress) Altered root branching and root morphology Norman et al. 1996; Fusconi et al. 1999 Hormonal changes (e.g., abscisic acid, auxins, cytokinins, ethylene) Allen et al. 1980, 1982; Danneberg et al. 1993; Dugassa et al. 1996; Hirsch et al. 1997; Torelli et al. 2000 Damage compensation Cordier et al. 1996; Pozo et al. 2002a Biochemical changes associated with plant defence mechanisms and induced resistance Phenolics and phytoalexins Morandi 1996 Amino acid levels (e.g., arginine, proline) Baltruschat and Schönbeck 1972; Giovannetti et al. 1991 Internal structural barriers (e.g., lignins, callose, hydroxyproline-rich Benhamou et al. 1994; Matsubara et al. 1995; Cordier et al. 1998; Pozo et al. glycoproteins) 2002b Defence-related proteins (e.g., pathogenesis-related proteins, β-1,3- Dugassa et al. 1996; Morandi 1996; Pozo et al. 1996, 1998, 1999, 2002a, glucanases, 2002b; chitosanases, chitinases, peroxidases, phenylalanine ammonia lyase, chalcone Mohr et al. 1998; Slezack et al. 2000; Guillon et al. 2002 synthase, superoxide dismutase) Increased DNA methylation and respiration Dugassa et al. 1996 Systemic induced resistance Cordier et al. 1996, 1998; Pozo et al. 2002b Development of an antagonistic microbiota Bacteria and fungi Meyer and Linderman 1986a; Secilia and Bagyaraj 1987; Thomas et al. 1994; Citernesi et al. 1996; St-Arnaud et al. 1995,1997; Andrade et al. 1997, 1998; Hodge 2000; Vázquez et al. 2000; Filion et al. 2003 1.2.3. Effect on soil structure Soil structure depends on the type and size of aggregates, and is an important component of the nutrient cycling system. The AMF (roots and extra-radical hyphae) act as driving factors for macroaggregates (>250 µm) stability, while microbial and plant residues, polysaccharides produced by bacteria and inorganic materials stabilize

Chapter 1 GENERAL INTRODUCTION microaggregates (<250 µm) (Tisdall, 1994). The formation of macroaggregates can prevent wind and water erosion, and both the quality and the size distribution of soil aggregates can affect soil porosity. Together, these properties can influence soil physical, chemical and biological processes, through effects on the accessibility of carbon, shelter, water, oxygen and nutrients to soil biota (Miller and Jastrow, 1992). AMF are considered to be primary soil aggregators through direct or indirect effects. The direct effects involve the AMF acting as binding agents. A positive relationship between AMF soil hyphae (some can grow as long as 90 mm, in Camel et al., 1991) and aggregate stabilization has been identified (Andrade et al., 1998; Jastrow et al., 1998). Indirect effects on water soluble aggregate (WSA) stabilization were described by Rillig et al. (1999), via the production of glomalin-related soil protein (GRSP). However, the underlying mechanisms associated with this system are not fully understood. The percentage of WSA depends on root and AMF hyphal lengths and also on interspecfic interactions (Piotrowski et al., 2004). 1.3. COST-BENEFIT OF THE SYMBIOSIS The cost-benefit analysis can be used to evaluate the benefits provided by mycorrhizal associations (enhanced mineral nutrients uptake) against the costs (carbon supplied by the host). Arbuscular mycorrhizal plants have been estimated to allocate 4-20% more photosynthates to roots compared to non-mycorrhizal plants (Jakobsen and Rosendahl, 1990; Smith and Read, 1997). The growth of the obligate biotrophic fungi in roots relies on carbon transferred across interfaces between the plant and the fungus, which means that the symbiosis is always obligatory for the AMF, whereas for the individual plant it is usually facultative. Some plants, however, can fully depend on the association, at least at some soil fertility levels (Brundrett, 1991). 1.3.1. Plant mycorrhizal dependence vs. plant responsiveness Mycorrhizal dependence should not be confused with host plant responsiveness to mycorrhiza as was explained by Janos (1988). Mycorrhizal dependence measurements are intended to accurately assess the need of plant to form mycorrhizal associations. Host plant responsiveness to mycorrhiza, which is used as a measure of fungal effectiveness, is influenced by the host plant, AM fungus and soil fertility level. The plant response to 21

Chapter 1- MYCORRHIZAE AN OVERVIEW inoculation changes if different AM fungal species are used as inoculants, or if edaphic conditions are altered (Sieverding, 1991). 1.3.2. Mycorrhiza association - from mutualism to parasitism For the majority of plants, the benefits of exchanging photosynthates for mineral nutrients outweigh the costs. However, this trade of nutrients is not always mutualistic (i.e., beneficial for both symbionts), and under some circumstances the net cost of the symbiosis exceeds the net benefit (Francis and Read, 1995). The full spectrum of plant responses to the formation of mycorrhizae can range from mutualism to parasitism (Johnson et al., 1997; Klironomos, 2003; Jones and Smith, 2004). Several factors can determine that variation, just as plant taxa vary in mycorrhizal dependency, and fungal taxa vary in mycorrhizal effectiveness. However, it cannot be concluded that a certain AMF genotype is always parasitic, because the cost-benefit balance in a pair of host plant- AMF species can change over time according to edaphic or seasonal factors (Fig. 1.2). Lapointe and Molard (1997) demonstrated that plant benefit from the presence of mycorrhizae varies during the life cycle of Erythronium americanum: in fall, mycorrhizae were more costly for the plant in terms of carbohydrate reserves. Figure 1.2. Mycorrhizal phenotypes are manifestations of the interaction between plant and fungal genotypes and environmental conditions. These factors determine the functioning of mycorrhizae along the mutualism-parasitism continuum (in Johnson et al., 1997).

Chapter 1 GENERAL INTRODUCTION In the extreme situation of the mycorrhizal association, where the plant or fungus become parasitic, the partner is then defined, in ecological terms, as a cheater (Janos 1987; Smith and Smith, 1996; Martins-Loução, 2002). A cheater in a symbiosis is the individual that receives the benefits of mutualism but does not reciprocate (Soberon and Martinez del Rio, 1985). The explanation for this occurrence remains unclear and is probably related to the outcome of the symbiosis. The host fungus interface of mycorrhizal associations may have evolved, in part, as a mechanism to limit cheating by tightly coupling the costs and benefits of the exchange for both partners. This process seems to be primarily controlled by the plants through different mechanisms, which can escape from their obligations in the associations (facultative) (Smith and Smith, 1996; Pinior et al., 1999; Redman et al., 2001; Brundrett, 2002). Plant mechanisms for preventing unwanted colonization may not be specific enough to distinguish cheaters from beneficial fungi. The most effective mechanism for plants to stop fungi from absorbing photosynthates without providing benefits is by suppressing root colonization, particularly by arbuscules. Klironomos et al. (1993) found that hyphal coils were prevalent but arbuscules were rare in maple forests. Pawlowska et al. (1996) reported the mycorrhizal status of plant species, where AMF fungi were present in the roots forming vesicles, but no arbuscules were found. 1.3.3. Preferential associations between host plant-amf species In 1983, Hayman demonstrated that certain fungi can be more effective than others to stimulate plant growth, or have no influence at all. Other studies also revealed such a wide range of outcomes (e.g., Gianinazzi and Gianinazzi-Perason, 1986; Bethlenfalvay et al., 1989; McGonigle and Fitter, 1990; Sylvia et al., 1993; Wilson and Hartnett, 1998; Graham and Abbot, 2000; Klironomos, 2003). Recently, Klironomos et al. (2000) demonstrated that two AMF species (G. etunicatum and G. intraradices) induced plant biomass differently, and suggested that those responses were due to differences in the carbon cost for plants maintaining the symbiosis with different fungi. Those differences can be related to variation in intra- and inter-radical growth rates of each AMF species (Hart and Reader, 2002a,b) and hence, to differing needs for plant carbon, or to different abilities to obtain it. In addition, different AMF vary greatly in the amount of phosphorous transported to the plant, and plant identity is an important determinant of the amount of phosphate transported (Jakobsen et al., 2002). Conversely, the effects of a change in photosynthate availability in the plant due to herbivory affect the mycorrhizal relationship differentially 23

Chapter 1- MYCORRHIZAE AN OVERVIEW depending on the species of fungus involved or the AMF community composition (Klironomos et al., 2004) The differential gains that plants receive from a mycorrhizal association in relation to AMF taxa, can be found either at family level (Hart and Reader, 2002a) or at species level (Sanders and Fitter, 1992; Bever et al., 1996; Klironomos, 2003). These differences in the outcome of the mycorrhizal symbiosis were recently interpreted based on functional compatibility and diversity concepts (Ravnskov and Jakobsen, 1995; Klironomos, 2000), and even on the presence of some level of specifity (Sanders, 2002, 2003). Even though the mycorrhizal association has been described as non-specific, given the AMF species ubiquitous distribution and the capability to infect a wide range of plants, there is evidence of host-specificity. Results from recent studies have revealed a differential host plant-amf species combination, which may imply some preference or even specificity among arbuscular mycorrhizal associations (McGonigle and Fitter, 1990; Bever, 2002). Nevertheless, it has not been shown whether plant or AMF species are capable of selecting a suitable partner. By definition, specificity implies compatibility between two organisms, with molecular recognition between them. For arbuscular mycorrhizae this specificity concept has to be differently interpreted, considering the genotype (of both symbionts) and the environmental features involved (Sanders, 2002). 1.4. ROLE OF MYCORRHIZAE IN MAINTAINING PLANT COMMUNITY DIVERSITY, LINKING ABOVE AND BELOW GROUND Since a single plant can form mycorrhizae with many fungi (Tommerup, 1988; Helgason et al., 1998), and a single fungus can connect many plants (Allen, 1996), the number of possible combinations of symbionts is significant, and so is the variety of possible outcomes from the symbiosis. This fact, associated with the patchy distribution of AMF species within communities (Klironomos et al., 1993, 1999; Carvalho et al., 2003), may play a role in determining plant community composition. 1.4.1. Effect of AMF communities on plant communities The presence vs. absence of AMF (e.g.: Grime et al., 1987; Gange et al., 1993; Hartnett and Wilson, 1999; O Connor et al., 2002) and AMF community composition (e.g., Streitwolf-Engel et al., 1997; van der Heijden et al., 1998a, b; Kiers et al., 2000) can greatly influence the structure and productivity of plant communities. Grime et al. (1987)

Chapter 1 GENERAL INTRODUCTION and Gange et al. (1993) showed a positive relation between mycorrhizal and plant diversity, i.e. the presence of AMF would increase plant diversity. However, Hartnett and Wilson (1999) and O Connor et al. (2002) showed an inverse relation between mycorrhiza and plant diversity. The positive relation between AM and plant diversity may be due to an increase in plant eveness through the transfer of assimilates from dominat species to the other species via the external mycelium. Hartnett and Wilson (1999) proposed that the differential host species response to fungal colonization limits plant species evenness, when mycorrhizal plant species become dominant, and no changes in the total biomass of plant community occur. O Connor et al. (2002) observed in a semiarid herbland, that the behavior of individual dominant mycorrhized plant species, was a strong determinant of community structure. The suppression of mycorrhizal activity with equivalent mycorrhizal responsiveness but different capacities to exploit the symbiosis for nutrients uptake or alleviation of water stress, resulted in changes in plant competition and redistribution of plant biomass within the community. In addition, studies at the AMF community level showed that, not only the presence or absence of AMF inoculum, but also the diversity and identity of AMF induced differential effects in plant growth, being determinant of plant diversity and productivity (van der Heijden et al., 1998). These authors also found that plant biomass increased with increasing AMF diversity, although productivity saturated after a certain level of AMF species diversity was reached. So these results support the redundancy hypothesis (see Hart et al., 2001). It has been suggested that the underlying mechanisms by which AMF promote growth of such mycorrhizal dependent plant species can be used to explain how AMF promote plant diversity (van der Heijden, 2002). However, the mechanisms are not yet explicit. 1.4.2. Effect of host plants on AMF communities AMF community composition can be influenced by host plant species (e.g., Johnson et al., 1991; Bever et al., 1996; Bever, 2002), by plant community composition (Eom et al., 2000; Burrows and Pfleger, 2002; Lovelock et al., 2003), and by soil factors such as soil moisture gradient (e.g., Anderson et al., 1984), fertility (Johnson, 1993), namely N concentrations (Egerton-Wharburton and Allen, 2000), textural differences (Allen et al., 1998), and ph (Moutoglis and Widden, 1996). Host plants may be one of the most important factors regulating AMF species composition, mainly through carbon allocation to roots, since every phase in AMF life history (colonization, hyphal spread in the roots and sporulation) is directly influenced by plant roots. Herbaceous plant species were 25

Chapter 1- MYCORRHIZAE AN OVERVIEW shown to induce changes in AMF community composition through significant differences in species richness and evenness (Eom et al., 2000). While tropical tree species were shown to induce only changes in relative abundance of a few common AMF species (Lovelock et al., 2003). These findings suggest that roots from different plant species offer different environments for reproduction, growth or survival of different AMF species or genera. 1.4.3. Host plant-amf interactions: feedback dynamics relation Bever et al. (e.g., 1997, 1999) suggested a feedback model as the underlying mechanism behind the plant-amf mutual interactions. Theoretically, a feedback response occurs between interacting organisms when one organism (A) affects the growth of another (B), which in turn has a positive or negative effect on the performance of the first organism (A); these relations can be self-promoting (positive feedback) or self-limiting by affecting the growth of other organism (negative feedback) (Bever et al., 1997; Hart et al., 2003) (Fig 1.3). Figure 1.3. Feedback due to changes in composition of the community of soil mutualists. The direction of benefit delivered between two plants species, A and B, and their fungal mutualists, X and Y, are indicated by arrows, with the thickness of the arrow indicating the magnitude of benefit. Positive feedback, the presence of plant A will result in an increase in the frequency of fungus X relative to fungus Y, which will then increase the rate of growth of plant A relative to plant B. Negative feedback, the presence of plant A will result in a decrease in the frequency of fungus X relative to fungus Y, which will then decrease the rate of growth of plant A relative to plant B. (in Bever et al., 1997). Exemplifying for the mycorrhizal association, positive feedback occurs when certain host plant species gradually select AMF assemblages that can optimize their growth, while negative feedback occurs when selected AMF have such heavy carbon demands that actually impair host growth. These findings have wider implications at community level. Through the positive feedback approach, the AMF community would increase the

Chapter 1 GENERAL INTRODUCTION growth rates of a particular host plant species that could become the dominant species, leading to a decline in plant species diversity. Through the negative feedback approach, the AMF community would decrease the relative growth rates of the most abundant plant species through a weaker mutualism relation, leading to the coexistence of competing plant species (Kiers et al., 2000; Klironomos, 2002; Bever, 2003). Additionally, the benefit that certain plant species receive from its AMF community can degrade over time, and this process leads to the dissolution of the mutualism and to an evolutionary feedback dynamics over time (Bever et al., 2002). The previous findings describing that differential host plant-amf species interactions may regulate each organism growth and fitness, and consequently induce shifts at plant- AMF community level, were recently extended to more complex interactions within plant- AMF-soil microorganisms. These interactions can influence plant community diversity and dynamics, by facilitation of invasive plant species or by mediating plant competition (van der Putten et al., 1993; Mills and Bever, 1998; Packer and Clay, 2000; Klironomos 2002; Callaway et al., 2004), enhancing the importance of mycorrhizal associations at the ecosystem level. Reynolds et al. (2003) suggested that positive feedback between plants and soil microbes plays a central role in early successional communities, while negative feedback contributes both to species replacements and to diversification in later successional communities. 1.5. EFFECT OF DISTURBANCES ON MYCORRHIZAE, PARTICULARLY IN MEDITERRANEAN ECOSYSTEMS Arbuscular mycorrhizae are vulnerable to disturbance and it has been frequently shown that disturbance causes a reduction in AMF community composition and functioning. A more detailed review of the available studies shows that there is large variation in the results of the effect of disturbance on AMF populations: some report a great loss of AMF viable propagules, spores (Miller et al., 1985; McGee, et al., 1997) and hyphal network (Jasper et al., 1989a,b; Evans and Miller, 1990; McGonigle and Miller, 1993), and their subsequent infectivity (Evans and Miller 1988; Jasper et al., 1989c; Abbott and Robson, 1991; Lovera and Cuenca, 1996; McGonigle and Miller, 1996); other studies only show a reduction in propagule number, and not in the AMF infectivity (Jasper et al., 1991; Miller et al., 1995, Gavito and Miller, 1998). This variability in AMF responses to disturbance is probably dependent on (i) the severity of the disturbance, as well as on (ii) the abiotic and biotic conditions prior to disturbance, (e.g,. soil nutrient 27

Chapter 1- MYCORRHIZAE AN OVERVIEW concentrations, the AMF community composition and the successional stage of the plant community composition) (Miller and Jastrow, 1992; McGonigle and Miller, 1993). i) Disturbance severity is increased by frequency and intensity of the action (Abbott and Gazey, 1994). Examples of severely disturbed ecosystems may include those with recently experienced volcanic activity, as well as those used for the extraction of aggregates, such as quarries and gravel pits. After disturbance, soil resources may not be limiting, but AMF inoculum potential may be reduced or eliminated, limiting the establishment of mycotrophic plants (Hart and Klironomos, 2002). Research on Mt. St. Helen, after the 1980 eruption, showed that the landscape was devoid of AMF, and only after AMF inoculum was reintroduced into the system by the activities of animals, a wide variety of mycorrhizal-dependent plants began to establish (Allen, 1991). Surface mining is another example of the most severe disturbances; three degradation factors are inherent to this activity: removal of growing plants, soil disturbance and topsoil storage. Despite the large impact of disturbance caused by heavy machinery (destroying plant roots and soil structure), stockpiling the soil is the most detrimental process. Stockpiling causes a negative effect on soil aggregation, litter decomposition and microorganisms diversity and survival, including AMF (Waaland and Allen, 1987; Stahl et al., 1988). Although mycelium, their living biomass is destroyed, AMF are able to resist to the most extreme condition, since spores can remain dormant for a long time. In stockpiled soil, the inoculum potential decreases because top soil is continuously diluted with deeper profile soil, which is poorer in AMF infective propagules (Nehl et al., 1999). However, the major negative effect in AMF infectivity is due to the reduction of extraradical hyphal network, which is very sensitive to physical disturbance. Disturbance in general, and soil stockpiling in particular, has different effects on different AMF species inducing shifts in AMF community composition (Abbott and Robson, 1991). This is probably due to different rates of sporulation within AMF species (Klironomos and Hart, 2002). For example, Visser et al. (1984) show that disturbance induced a change in the dominant fungi species, from G. fasciculatum to G. mossae. (ii) Abiotic and biotic components prior to disturbance. Edaphic parameters can also influence the effect of disturbance. For example, the stockpiled storage in semiarid conditions can reduce the AMF inoculum potential more than in moister environments. According to the hypothesis proposed by Jasper et al. (1991), AM infectivity is less affected by soil disturbance if there was high inoculum density prior to disturbance. However, McGonigle and Miller (2000) reported that high inoculum density did not prevent the negative effect of disturbance on AMF colonization. They suggested that if

Chapter 1 GENERAL INTRODUCTION those colonization differences appeared following soil disturbance they were probably dependent on the interactions between the environment and plants species. Mediterranean plant communities are continually subject to natural stresses: lower temperature when there is more moisture in the soil, and drier wheather during the growing season, together with low nutrient and water availabilities, and heterogeneous spatial distribution of AMF inoculum and vegetation (Carvalho et al., 2003). In addition, Stahl and Smith (1984) suggested that AMF species diversity and spore density decreased with increased aridity in semi-arid grasslands, which agrees with other studies reporting low AMF propagule density in semi-arid areas (e.g. McGee, 1989; Requena et al., 1996; Ferrol et al., 2003). Stutz and Morton (1996) contradicted these findings demonstrating that AMF diversity in a semi-arid region was similar to the one found in other plant communities after three cycles of trap culturing. Nevertheless, the spore density found in this study cannot be related to species abundance in the field. In addition, mycelium is the main infective propagule in Mediterranean ecosystems (Requena et al., 1996; Bashan et al., 2000) and remains infective after the Mediterranean natural cycles of soil dryness (Jasper et al., 1987, 1993; Braunberger et al., 1996). Together, these findings support the idea that AMF communities are vulnerable to soil disturbance in these ecosystems comparatively to others. During ecosystem recovery from disturbance, changes take place in plant populations (secondary succession). There is evidence that in some situations mycorrhizal inoculum may be a factor influencing the rate of succession. During this process, non-mycorrhizal plants are gradually replaced by obligate mycorrhizal plants later in succession (Miller, 1987; Allen, 1991; Francis and Read, 1994; Barni and Siniscalco, 2000), and the AMF spore abundance and species richness increase along the successional gradient (Allen, 1991; Johnson et al., 1991; Gemma and Koske, 1997). In semiarid ecosystems, where natural secondary succession is very slow (Bradshaw, 1983), the influence of mycorrhiza in controlling plant diversity may be vitally important. Plant communities can be subject to dramatic seasonal and interannual fluctuations, and may rely on AMF high biodiversity to maintain stability (Grime, 1997). 1.6. MYCORRHIZAL ROLE IN THE RESTORATION OF DEGRADED MEDITERRANEAN ECOSYSTEMS Several studies performing large scale field restoration assays demonstrated that manipulation of AMF community guarantee the success of plant establishment, 29

Chapter 1- MYCORRHIZAE AN OVERVIEW overcoming AMF propagule limiting factor (Cuenca et al., 1998; Dodd et al., 2002; Caravaca et al., 2004). Through appropriate AMF inoculum management in the restoration process, it is possible to recover not only the vegetation but also the biological and physico-chemical soil properties (Miller and Jastrow, 2000; Jeffries and Barea, 2001, Jeffries et al., 2003; Caravaca et al., 2005). Reeves et al. (1979) and Janos (1980) had already suggested that increasing AMF population would possibly overcome the negative effect of disturbance on AMF, and therefore accelerate natural secondary succession. However, Miller and Jastrow (1992) alert to the fact that AMF manipulation would not facilitate the skipping of successional stages, but would rather prevent the stagnation of community development. In the early 1990 s some protocols appeared to ameliorate the disturbance effect, suggesting AMF inoculum management for restoration (Dodd and Thomson, 1994; Jasper, 1994) and agricultural purposes (Sieverding, 1991). All of them began with an evaluation of indigenous AMF infectivity in disturbed soil, as well as a test of AMF effectiveness. If both evaluation yielded poor results, different AMF inoculum manipulations were suggested, but the most recommended one was the addition of AMF inoculum. Sieverding (1991) and Jasper (1994) suggested two manipulation strategies: (i) manipulation of indigenous AMF communities or (ii) soil inoculation with selected AMF species (Fig. 1.5). i) through indirect selective practices, like using appropriate cropping plant species, it is possible to build up the indigenous inoculum. This approach is recommended when indigenous AMF infectivity is low but with adequate effectivity. The advantages of this strategy are related to the presumed adaptation of these AMF to the native plants and to environmental conditions. Nevertheless, if only AMF propagule density is increased, no changes will occur in the diversity of the disturbed AMF community, and the existing AMF community may or may not be effective in promoting plant establishment and growth. ii) through direct soil inoculation with allochthonous AMF species. Inoculation would imply a significant change in indigenous AMF community, increasing the diversity and density of AMF species in disturbed soil. Dodd and Thomson (1994) proposed a protocol to identify the responsive sites for mycorrhizal inoculation, accompanied by another protocol for isolating and screening the most effective mycorrhizal species/isolate, in order to artificially boost the indigenous

Chapter 1 GENERAL INTRODUCTION inoculum by using a range of baiting and plant culture techniques. Other authors agree on the importance of considering the sources of mycorrhizal fungi in any screening and selection trail. Also, it may be ideal to specifically isolate candidate organisms from target sites such as those of future use (Abbott and Gazey, 1994; Requena et al., 1997, 2001; Lovato et al., 1999). However, none of the AMF management protocols suggests to test the commercial inoculum effectivity on native plants prior to restoration programs, nor the evaluation of the effects on the remaining AMF community composition after inoculation. Figure 1.4. A schematic representation of a protocol for managing mycorrhizal fungi in soils to be revegetated. (in Jasper, 1994). If AMF inoculum is missing in the disturbed soil, the addition of commercially available inoculum is suggested. The basic criteria for selection and characterization of AMF inoculants is the ability to achieve improved plant growth (Abbott and Gazey, 1994; Estaún et al., 2002; Vosátka and Dodd, 2002; Gianinazzi and Vosátka, 2004) and/or that they act as agents for disease control (Smith and Read, 1997; Varma and Hock, 1999; Barea et al., 2002; Gianinazzi and Vosátka, 2004). AMF can be considered as a biocontrol agent because it was shown that AMF reduced the detrimental effects of pathogens in plants beyond the additive effects resulting from an improved plant nutrition (Borowicz, 2001). Sylvia et al. (1993), in a complex experiment involving wide range of soils (with different levels of P content), AMF isolates and two host plant species, showed that two Glomus isolates were effective for the majority of soils. These findings were relevant for those involved in producing AMF inoculum at commercial scale, demonstrating that AMF could adapt to different edaphic conditions, and that it is possible to select effective isolates. However, the strategy of adding commercial inoculum has been over-generalized and suggested for application, independently of the ecosystem type and disturbance severity, without being pre-tested for effectiveness in the target plants. In general, it is assumed that regardless of their diversity, AMF species act as homogeneous functional 31

Chapter 1- MYCORRHIZAE AN OVERVIEW groups in promoting plant benefit for any host-plant species. Recent studies on mycorrhizal ecology have shown that this assumption is not verified in all cases. Different AMF-plant host pair combinations have different outcomes in symbiosis effectivity. Gianinazzi and Vosátka (2004) suggested strong collaboration between inoculum producers, plant growers and researchers to improve the understanding of AMF ecophysiology, in order to achieve an efficient AMF manipulation, since it is difficult to predict the symbiosis outcome in terms of plant growth and health. Little is known about the stability of plant growth response to commercially available AMF inoculum over a range of host plants species and soil environments. Some studies have been recently published concerning the improvement of plant growth in Mediterranean ecosystems through direct field inoculation with allochthonous AMF species (Herrera et al., 1993; Bhatia et al., 1998; Requena et al., 2001; Barni and Siniscalco, 2000; Caravaca et al., 2003a, 2005). Herrera et al. (1993) showed that AMF helped to establish native woody plants in a long-term Mediterranean revegetation study. The author emphasized the importance of mycorrhizal biotechnology for nursery of native plant species for revegetation programs, because the symbiosis enhances the ability of these plants to become established and cope with natural stresses in such ecosystems. Despite these allochthonous AMF species are usually selected for their extreme aggressiveness (see Graham and Abbot, 2000), their effectiveness in promoting native Mediterranean plants growth is sometimes unsuccessful, inoculation with indigenous AMF species proving to be more effective when plants are outplanted (Requena et al., 2001; Caravaca et al., 2003a; 2005; Alguacil et al., 2005;). In addition, the variation of inoculation effectiveness, regardless of the AMF inoculum source, is plant species dependent (Caravaca et al., 2003a) and the reciprocal is also verified, which means different native species promote diferential mycorrhizal propagule production (Palenzuela 2002; Azcon-Aguillar et al., 2003; Caravaca et al., 2003b). Other studies in Mediterranean ecosystems that involved soil inoculation did not find a positive relation between the increment of soil AMF propagules through inoculation and better plant establishment or plant productivity, either using AMF inoculum from greenhouse cultures (Richter and Stutz, 2002; Bell et al., 2003) or commercial inoculum (Werner et al., 2001; Clemente et al., 2004). Nevertheless, the scarce references in literature referring the effects of AMF commercial inoculum on woody Mediterranean plant growth, and on indigenous AMF community, are even more limited in the case of commercial inoculum abilities to compete for Mediterranean woody plant roots with non mycorrhizal fungi.

Chapter 1 GENERAL INTRODUCTION 2. OBJECTIVES AND RATIONALE OF THE RESEARCH AMF are important elements of terrestrial ecosystems, and thus considered as key components for a successful restoration, linking above and below ground communities. However, inconsistent results in plant growth benefit have been documented concerning the application of AMF soil inoculum, ranging from positive (see Lekberg and Koide, 2005) to null (eg: Bell et al., 2003) and to detrimental outcomes from the symbiosis (eg: Modjo and Hendrix, 1986; Hendrix et al., 1992). Nevertheless, the fundamental concept of AMF inoculum management approaches relies on the assumption that there is a positive linear relationship between the presence and/or abundance of AMF soil propagules and plant productivity, and that AMF are functionally redundant in their effects on a host, not considering the AMF species diversity effect and the complexity of different interactions of AMF with plant species, microbial communities and soil abiotic characteristics. Little is known about the possible underlying mechanisms behind the inconsistent results of inoculation. Considering recent scientific concepts like the relation between AMF diversity and plant productivity, the mycorrhizal functioning in the mutualismparasitism continuum, the evidence of functional compatibility between pairs of host plant- AMF species, and the plant-amf feedback interaction the major aims of this thesis were: i) to provide more detailed information about the effects of introduced AMF species, as commercial inoculum, on plant growth of different Mediterranean woody plant species; ii) to improve the understanding of the role of introduced AMF species as mediators between the Mediterranean woody plants growth and indigenous AMF community diversity and functioning. Specifically, the following questions were addressed: 1. Does disturbance act as a selective force, shifting the AMF community composition by selecting less infective and effective AMF species? (Chapter 2) 2. Do changes in species diversity and abundance of AMF communities, driven by disturbance and commercial inoculum addition, induce changes on native plant growth benefits? Are those changes plant-species dependent? (Chapter 3) 33

Chapter 1- OBJECTIVES AND RATIONALE OF THE RESEARCH 3. Do changes in AMF communities, driven by disturbance and commercial inoculum addition, induce changes on plant-root endophyte interactions? (Chapter 4) Since the essential component of this work is the evaluation of the effects of AMF inoculum application and the interpretation of the underlying mechanism of its action, a situation where the use of commercial inoculum is undoubtedly recommended was selected (Jasper, 1994; Requena et al., 1996, 2001; Caravaca et al., 2004): the restoration of a very disturbed site in Mediterranean ecosystems. An innovating experimental system was set up to achieve the former goals and to obtain more conclusive results from the comparison of different AMF community functions. The use of removal experiments (with or without AMF) and the manipulation of AMF species, building artificial communities, were avoided. The control of other variables which may also induce plant growth differences, such as soil nutrient levels and soil microbial community, was also sought. More detailed information on how these conditions were achieved is given in Materials and Methods of Chapters 3 and 4. In this study, different AMF communities were used: i) the AMF community from a disturbed soil, and ii) the AMF community from a disturbed soil plus that from an undisturbed soil. Comparing these two AMF communities it would be possible to infer the outcome of the addition of commercial inoculum by itself. The effects of those communities on plant growth were further compared with those of the AMF community in an undisturbed soil, retrieved from a surrounding well-preserved scrub. To compare the different AMF community functions, plant biomass, root colonization and AMF soil inoculum were consistently used as key estimators of symbiosis effectiveness for plant and AMF performance. Root colonization represents the contact interface between the two symbionts. For a more accurate assessment of symbiosis functionality, internal mycelial structures - arbuscules, vesicles, and hyphae - were also evaluated, following previous recommendations (Hart and Reader, 2002b). Underlying hypotheses and expected results In Chapter 2, different methodological approaches were used to test whether disturbance has a selective effect on AMF inoculum and on plant growth. It was expected that disturbance acted as a selective force on the AMF community, reducing propagules density and diversity, comparatively to the AMF community from the undisturbed soil.

Chapter 1 GENERAL INTRODUCTION Consequently, root colonization by AMF would be reduced (McGonigle and Miller, 1996) (Fig. 1.5). In addition, the growth benefit for individual plants would decrease (Fig. 1.5), although it was expected that the rate of benefit reduction would depend on the host plant species. Given that AMF differ in their symbiotic function and that plant species differ in their responses to AMF, then communities with higher AMF species richness could mean more functions fulfilled and more opportunities for beneficial relationships to develop. Since disturbance may induce a shift in AMF community, reducing diversity, it is hypothesized that plants would grow better in the presence of the AMF community from the undisturbed soil. Moreover, the tested plants were from a middle - late successional stage, and according to Allen (1991) and Johnson et al. (1991) it would be expected that these plants would grow better with late successional AMF communities. Plant biomass % root colonization AMF soil inoculum Disturbance effect Figure 1.5. Schematic representation of hypothetical disturbance effect for the dependent variables: plant biomass (---), AMF soil inoculum and percentage root colonization ( ). The hypothesized variability of AMF variables starts with a plateau, corresponding to the well established AMF soil community in the undisturbed soil and to the root colonization of native plants living in it. Chapter 3 presents the experiment designed to test whether inoculum of AMF communities from a disturbed soil with or without AMF commercial inoculation, in comparison with AMF communities from undisturbed area, led to different host plant growth. The effects of plant species on AMF soil inoculum was also tested. For the hypothesis formulation, two approaches are possible: i) The concept that the addition of AMF commercial inoculum to promote plant growth depends only on AMF soil propagules density, assumes that AMF act as a homogeneous functional group, regardless of their species diversity. This means that biodiversity is assumed to be functionally redundant even with a low number of 35

Chapter 1- OBJECTIVES AND RATIONALE OF THE RESEARCH species. Therefore, in this thesis it is hypothesized that target plant species colonized by the AMF community from the undisturbed native area would present similar growth as compared to plants colonized by the AMF community from the disturbed soil previously inoculated with commercial inoculum (Fig 1.6). (ii) If there is functional compatibility between pairs of plant-amf species and if AMF community diversity influences plant productivity, then a differential plant growth among the three tested woody species is expected, relatively to the AMF communities treatments. Plant biomass Time Figure 1.6. Schematic representation of hypothetical plant growth responsiveness to three different AMF communities: AMF community from the undisturbed soil, AMF community from the disturbed soil inoculated with commercial inoculum, AMF community from the disturbed soil. It is assumed that AMF biodiversity is functionally redundant. It is hypothesized that each plant species responsiveness to AMF activity (sensu Wilson and Hartnett, 1998) would be largely variable since each community can be composed of several species. Therefore, it is possible that some mutualistic preferential association occurs between native plant species and certain adapted indigenous AMF species; this would be revealed as a reciprocal benefit for both symbionts measured as an increased plant growth and inoculum potential, although the rate of inoculum build-up could be plant species-specific. The beneficial plant response of the addition of commercial inoculum would be plant species-specific. In Chapter 4, it was tested whether individual plant species induce shifts in soil fungal communities (conspecific vs. heterospecific communities), and whether plant growth is increased or decreased by these communities. As part of this objective, the possible capacity of AMF species from commercial inoculum to alter the feedback mechanism of plant-root fungal endophyte interactions was also investigated.

Chapter 1 GENERAL INTRODUCTION It was expected that the addition of commercial inoculum would reduce the negative feedback among the interactions of conspecific plant-root endophyte community. Since AMF species from commercial inoculum are usually previously selected for their aggressive infectivity abilities, it is hypothesized that they would act as a biocontrol agent, either by increasing competitiveness for root space against other root endophyte fungi, or by inducing plant physiological changes and morphological root differences. Since not all AMF species exert the protection function similarly, it was expected that plant growth increased as a consequence of pathogens infection amelioration, and changes would occur differentially for each plant species according to AMF communities and non-mycorrhizal root fungi species diversity. 3. EXPERIMENTAL FRAMEWORK Mediterranean climate is characterized by dry and hot summers contrasting with wet and relatively cold winters. Mediterranean-type ecosystems correspond to a transitional region between temperate and tropical ecosystems and have a restricted distribution, (Mediterranean Basin, California, Chile, South Africa and some southern areas in Australia) (Di Castri et al., 1981). The complex biological diversity of these ecosystems is a result of a recent evolution and adaptation to climatic changes, together with several selective driving forces: biogeography, geology, ecology and history (Blondel and Aronson, 1999). The historical patterns of resources management in the Mediterranean Basin often resulted in their overexploitation, with profound impact on structure and composition of the vegetation. The exceptional richness of plant species is partially a result of long-standing human activities (fire-setting, clear-cutting, grazing by domestic livestock, etc.) (Naveh, 1975; Blondel and Aronson, 1999; Vallejo et al., 2006). Ecosystem heterogeneity ( mosaic effect ) caused by the combination of those factors acts not only as a consequence but also as a main evolutionary factor for Mediterranean species. Abiotic (e.g. water, nutrients) and biotic (e.g. plant and microbial distributions), variables that typically have a patchy distribution in these ecosystems (e.g. Joffre and Rambal, 1993; Cruz et al., 2002; Carvalho et al., 2003). Generally, vegetation is a dense scrub (named maquis in the Mediterranean Basin) dominated by woody evergreen shrubs with sclerophyllous leaves, and summer semideciduous shrubs. There may be an overstory of small trees, as well as an understory of annuals and herbaceous perennials. Sclerophyllous shrubs are typical of middle-late successional stages. They present deep root systems and leaves with low surface-to- 37

Chapter 1- EXPERIMENTAL FRAMEWORK volume ratio and thick cuticles, features that favor the control of transpiration and are considered adaptations to summer drought and/ or to low soil nutrient availability, as is the case in Mediterranean-type ecosystems (Correia and Catarino, 1994; Blondel and Aronson, 1999; Werner, 2000). Semi-deciduous shrubs dominate in the earlier successional stages. They present shallow root systems and loose part of their leaves before summer drought, thus reducing their transpirational surface during that season (Correia et al., 1987, 1992; Kyparissis & Manetas, 1993; Werner, 2000). The extraordinary plant diversity (25 000 plant species) found in the Mediterranean Basin (Quézel et al., 1999) is accompanied by a high diversity of mycorrhizal types. The maquis comprises plant species able to form almost all types of mycorrhizae (Brundrett, 1991; Puppi and Tartaglini, 1991). Examples among woody species are Quercus coccifera (ectomycorrizal mycorrhizae), Arbustus unedo (arbutoid mycorrhizae), Erica arborea (ericoid mycorrhizae); the most common are the species establishing arbuscular mycorrizae such as Pistacia lentiscus and Olea europaea var. sylvestris. The above ground spatial complexity of this community is increased by the presence of different vegetation layers (trees, shrubs and herbaceous), creating a plant root mixed system from different plants species hosting different mycorrhizal fungi. Selected sites and plants Part of the soil samples (disturbed soil) used throughout the study were retrieved from a limestone quarry at Outão (SECIL company) where revegetation projects have been taking place since 1983. It is one of the largest limestone quarries in the region and is located within the Arrábida Natural Park, in Portugal (38º 29 46 N, 8º 57 00 W). The reference area from where undisturbed soil was collected - was adjacent to the quarry and is representative of a late seral native Mediterranean plant community (Fig. 1.7), suffering only from minor disturbance from peripheral mining activities (e.g. dust). Serra da Arrábida (38º 27-30 N, 8º 55-9º 02 W) is a small chain of limestone outcrops with a maximum elevation of 500m, which receives official protection since the 18th century, and is a Natural Park since 1976. The soils are classified as Mediterranean red soils, established on a Jurassic limestone. A true soil profile is almost absent. The natural vegetation in this area consists of a well-preserved Mediterranean maquis, dominated by evergreen sclerophyllous shrubs and summer semi-deciduous shrubs, respectively in later and earlier successional stages (Catarino et al., 1982; Correia,

Chapter 1 GENERAL INTRODUCTION 1988; Werner, 2000; Clemente, 2002). These plants are adapted to drought stress and lownutrient availability (Correia, 1988; Werner, 2000). The climate is Mediterranean, with a mean annual rainfall of 650 mm and mean annual temperature of 16ºC at the closest meteorological station (Setúbal) (data from Correia and Catarino, 1994). Soil is commonly denominated as terra rossa (red soils) and contains a large proportion of hard limestone and other calcareous rocks, mostly shallow. Generally, soils are poor in nutrients and have a neutral to basic ph (Correia, 1988). C A D E B Figure 1.7. Aspect of the reference undisturbed site, in Serra da Arrábida (A), of the greenhouse facilities where the experiments were carried out, in Praias do Sado (B), and plant species used carob (C), myrtle (D) and lavender (E). The experimental studies were carried house in the nursery facilities of the company EDP, at Praias do Sado, about 10 km from Serra da Arrábida (Fig. 1.7). Plants and seeds of Serra da Arrábida provenance, as well as maintenace of the growing plants, were provided by this nursery. For the purposes of this thesis, the selected native Mediterranean plant 39